CLIC DB injector study
-
Upload
zeus-davis -
Category
Documents
-
view
50 -
download
0
description
Transcript of CLIC DB injector study
CLIC DB injector study
LCWS 2012, Arlington, Texas, USA, October 22th -26th ,2012
Steffen Döbert, BE-RF
Introduction
Parameters and difficulties
CLIC DB injector front end
components status
CLIC 0
Conclusions
PREBUNCHER
BUNCHER
SLIT
ACCELERATING CAVITY
53 MeV
6 MeV
26 MeV
SHB
SOLENOIDS
CLIC DB injector schematics
THERMIONIC GUN
Documented in the CDR; a scaled version of CTF3
Simona Bettoni, Alessandro Vivoli
CLIC DB injector specifications and
challengesParameter Nominal value Unit
Beam Energy 50 MeV
Pulse Length 140.3 / 243.7 ms / ns
Beam current 4.2 A
Bunch charge 8.4 nC
Number of bunches 70128
Total charge per pulse 590 mC
Bunch spacing 1.992 ns
Emittance at 50 MeV 100 mm mrad
Repetition rate 100 Hz
Energy spread at 50 MeV 1 % FWHM
Bunch length at 50 MeV 3 mm rms
Charge variation shot to shot 0.1 %
Charge flatness on flat top 0.1 %
Allowed satellite charge < 7 %
Allowed switching time 5 ns
Sub Harmonic Bunchers (SHBs)
Hamed Shaker
Phase coding
180 phase switch
Acceleration n0
Deflection n0 / 2
Sub-harmonic bunching n0 / 2
odd buckets
even bucket
s
RF deflector n0 / 2
Gap creation and combination
Gun SHB 1-2-3
PB Buncher Acc. Structures
IOTs, 500 MHz
Modulator-klystrons, 1 GHz, 15 MW
~ 140 keV ~ 12 MeV
Diagnostics
Main challenge of SHBs: Fast 180° phase flipping
capability, simulation indicate < 18 ns
switching time needed
Simulation results
Longitudinal phase space after sub-harmonic bunching
Same total gap-voltage but fewer cells (2 instead of 6)
35-39 KV needed
Parameter Unit Simulations CLIC
Energy
Bunch charge
Bunch length (rms)
Energy spread (rms)
Horizontal normalized emittance (rms)
Vertical normalized emittance (rms)
Satellites population
MeV
nC
mm
MeV
mm rad
mm rad
%
53.2
8.16
2.83
0.45 (@53 MeV)
32.9
28.7
4.9
8.4
3 (@ 50 MeV)
< 0.50 (@ 50 MeV)
100
100
As small as possible
Simulation results
Specifications can be fulfilled but still high satellite population and high losses in cleaning chicane
Some ideas to improve the satellites and total losses Beam loading and wake field effects to be studied
Some inconsistencies due to a lack of realistic rf-parameters (simulations have to be redone with new parameters)
CLIC DB front end,Post CDR Project
Gun, sub-harmonic bunching, bunching, three accelerating structures,5 long pulse klystrons and modulators, diagnostics
Gun SHB 1-2-3
PB Buncher Acc. Structures
IOTs, 500 MHz
Modulator-klystrons, 1 GHz, 15 MW
~ 140 keV ~ 12 MeV
Diagnostics
~ 3 MeV
What do we plan to do until 2016
Optimistic and rough planning
Task 2012 2013 2014 2015 2016Space needed prepare gun test f acility prepare Klystron test sand prepare injectot building injector building
Gun conceptunal design design and construction GUN test f acility GUN test f acility
SHB Buncher design fabrication testing
500 MHz power source specifi cation purchase testing
Buncher specifi cation design and purchase reception, low power test high power test
1 GHz structure specifi cation design and purchase reception low power test high power test
Diagnostis specifi cation, purchase I C in gun test
LLRF specifi cation fabrication+test ready for klystron test
1 GHz klystrons tender design at manufacturer fabrication of prototype Receive Klystron 1 Klystron 2
1 GHz Modulator tender R&D R&D Receive fi rst MDK MDK2
I njector integration,
vacuum, controls,
magnets, diagnostics on hold on hold design
Create a gun test facility to test the sourceand a high power test stand to test the klystron, modulator and rf structures
Gun options
1. CTF3 type gridded gun with a HV modulator, Modulator stability, grid survival, several SHB needed, can be purchased
2. Modulated anode gun, needs new design and simulation, Can’t be bought of the shelf, not sure if modulation fulfills requirements
3. RF modulated grid, IOT-type gun, very attractive but needs tests and simulation, dark
current issue, emittance ?, reliability, can be purchased, no SHB, no satellites, likely R&D needed
RF Modulate classical gun, typical pulse length 1 ns, could have very low satellites, can this deliver enough current, space charge problems
Sub-harmonic bunching system
Status: RF design existing, ready to start mechanical design and launch prototype or cold model
Power source: 500 MHz, ~100 kW, wide band (70 MHz) sources needed for fast phase switching. Started to discuss with industry.
Candidates: IOT, frequency and power available, bandwidth to be seen
tests are planned with an 800 MHz IOT for SPS
Solid state amplifier, bandwidth, power, cost ?
RF design of 1 GHz pre-buncher and travelling wave buncher existing
Magnetic coupling TW structure
Hamed Shaker
l
g
rb
rn
θ1
r1
t/2
rc
lc
θc
g 40 mm
rb 45 mm
rn 4 mm
θ1 25°
t (disk thickness) 15 mm
Frequency 499.75 MHz
l (for 108° phase advance per cell)
≈115.18 mm
r1 161.55 mm
rc 142 mm
lc 54 mm
θc 86°
Phase velocity 0.64c
For the known gap voltage and filling time our goal is to increase R/Q to reduce the input power.
τ=10nsV=36.5 KV
Phase flipping simulation with beam – 10 ns
Hamed Shaker
port 1port 2
Beam Entrance
Continues beam with 6A current
Excitation signal from port 1 with 80 KW peak power.
Phase flipping simulation with beam – 10 ns
Hamed Shaker
Excitation signal – port 1
Output signal – port 2
Output signal – port 1
≈ 12 ns
≈ 44 ns
Phase flipping simulation with beam – 10 ns
Hamed Shaker
44 49 54 59 64 690
5
10
15
20
25
30
35
40
45
Maximum energy gain vs. Time
Time (ns)
Ma
xim
um
en
erg
y g
ain
(K
eV
)
About 5 bunches will be missed in comparison
with about 120 bunches in each sub-pulse.
DB-accelerator structure
Input and output coupler design finishedCorrect match, input reflection < 30 dB.(red and green: two different geometries; red is final)
RF-design existing, next steps: mechanical design and prototype
Rolf Wegener
DB-accelerator structure
Rolf Wegener
17 17.5 18 18.5 19 19.5 200
2
4
6
cell no
1-et
a(n)
[%
]
f0= 1.000 GHz, BP Radius= 49.00 mm, mean(Pin)= 15.00 MW
i_vgr=894
i_vgr=998i_vgr=845
i_vgr=975
17 17.5 18 18.5 19 19.5 20243
244
245
cell no
t_fil
l [ns
]
2 4 6 8 10 12 14 16 18 200
2
4
6
cell no
vgr/
c0 [
%]
Parameters:
f = 999.5 MHz
Pin = 15 MW
RB = 49 mm
N = 19 cells
OD= 300 mm
L= 2.4 m
Tfill = 245 ns
h RF-Beam= 97.5 %
RF power sources 1 GHz high efficiency
klystronScenarios 2012 2013 2014 2015 2016 2017
Multi beam klystron (ILC-based, 65% efficiency)
Tender/Order Design Klystron 1 Klystron 2
Klystron 3 ?
Ultimate efficiency (>70%) based on Chiara Marrelli’s work Study Study ?
Ultimate effiency based on EuCard2
Start of Progam
EuCard study
EuCard study
EuCard study
Transfer to Industry
Development and Purchasing strategy:Launch tender to develop and purchase high efficiency klystrons in industry. Aim for multiple vendors. Klystrons could arrive from 2015 in line with modulator development
Plan to hold a workshop on high efficiency klystrons and modulators in spring 2013 (EuCard2 network on high efficiency rf sources). Seeking collaborations for this topic !
Tentative klystron parametersPARAMETER VALUE UNITS
RF Frequency
Bandwidth at -1dB
RF Power:
Peak Power
Average Power
RF Pulse width (at -3dB)
HV pulse width (at full width half height)
Repetition Rate
High Voltage applied to the cathode
Tolerable peak reverse voltage
Efficiency at peak power
RF gain at peak power
Perveance
Stability of RF output signal
0.5-1.0 of max. power and 0.75 -1.0 of max. cathode HV to be:
RF input vs output phase jitter [*]
RF amplitude jitter
Pulse failures (arcs etc.) during 14 hour continuous test period
Matching load, fundamental and 2nd harmonic
Radiation at 0.1m distance from klystron
Output waveguide type
999.5
tbd
> 18
135
150
165
50
tbd, 150 (max)
tbd
> 65
tbd, > 50 ?
tbd
±0.5 (max)
±1 (max)
< 1
tbd
< 1
WR975
MHz
MHz
MW
kW
μs
μs
Hz
kV
kV
%
dB
μA/V1.5
RF deg
%
vswr
μSv/h
CLIC studies & klystron modulators
specsModulator main specifications
Pulse voltage Vkn 150 kV
Pulse current Ikn 160 A
Peak power Pout 24 MW
Rise & fall times trise 3 μs
Flat-top lenght tflat 140 μs
Repetition rate Repr 50 Hz
Flat-top stability FTS 0.85 %
Pulse reproducibility PPR 10 ppm
trise
FTS
Vovs
tset tflat tfall
Vkn
Vuns
treset
Time [s]
Vo
ltag
e [
V]
Trep
ideal pulsereal pulse
preal
pidealpulse E
E
_
_Pulse efficiency definition
~300MW required for kly. mod.
CLIC Drive Beam klystron modulators R&D strategy
Approach:Develop and explore with collaboration partners technologies to meet the ultimate specification for CLIC with the goal to have two working prototypes in 2015-2016.First collaboration with ETH Zürich started, prototype in 2015
Demonstration goals for the injector front end
Demonstrate rf system at full pulse length and beam loading High efficiency klystron High efficiency and stable modulator Full loaded accelerating structure (validate technology)
Demonstrate beam quality and stability requirements for long pulse Current stability 0.1% Beam phase stability Emittance and energy and position jitter
Demonstrate electron source and phase coding Life time, reliability, routine operation
Demonstrate diagnostics suitable for long pulse and machine protection
DB Injector suitable for CLIC zero and CLIC
Photo injector option
Advantages
No satellites or tails, phase coding on the laser sideNo or less bunching needed, possibly better emittanceFlexible time structure (single bunches)
Concerns
Cathode lifetimeChallenging laser, peak and average powerIntensity stabilityMaintenance and operationVery little resource available for time being
Introduction to
100 m
TBA
DBA0.48 GeV, 4.2 A DL
CR2
CR1
Compression2 x 3 x 4
DB Turn around0.48 GeV, 101 A
6.5 GeV, 1.2 A
0.2 GeV, 101 A
CALIFES type injector0.25 GeV, 1.2 A
H. Braun, CLIC 2008 Workshop
Czero
20 % of the CLIC Drive Beam and 10 % of a CLIC decelerator
Motivation for CLICzero
Demonstrate nominal drive beam parameters (except the energy) Full combination scheme to 100 A, full pulse length in injector
Realize two beam acceleration with nominal hardware for a
significant length (~100 m)
Drives industrialization, needed to be ready for CLIC Significant size series production of cost and performance
driving hardware (46 two beam modules, 276 x-band structures, 138 PETS, 140 L-band klystrons, modulators and structures)
Most hardware reusable for CLIC
Could be a beam driven processing facility for the x-band structures
Drawbacks: Expensive Beam dynamics for combination might be more difficult due to lower
energy Does not address sufficiently emittance preservation and luminosity
issues
Conclusions
Unusual parameter space for an injector
Rough conceptual design for the injector exists, now we have to get really started
Plans for purchasing the key hardware items have been developed and need to be followed up now
We seeking collaborations for this work, anybody interested ?
End
Klystron specifications
FREQVklystr
onIklystro
n
V pulse
width
RF pulse
width
Peak RF
Power
Repetition
rate
Average
PowerGain
Efficiency
Waveguide
MHz kV A µs µs MW Hz kW dB %1300 115 132 1700 1500 10 10 150 47 65 WR 650
TH1802, ILC MBK klystron
FREQVklystr
onIklystro
n
V pulse
width
RF pulse
width
Peak RF
Power
Repetition
rate
Average
PowerGain
Efficiency
Waveguide
MHz kV A µs µs MW Hz kW dB % 999.52 150 15-20 50 113 70
CLIC DB klystron design goal, ~ 150kV voltage was assumed for time being
Features of CLICzero
Nominal CLIC DB injector
Nominal DB rf system (Klystron, Modulator, accelerating
structure)
Nominal 100 A drive beam (6 ms)
20 % drive beam energy Nominal Delay Loop and Combiner Rings (1/5 of the energy)
Drive beam pulse shaping can be studied
DB turn around to study phase feed forward
46 nominal two beam modules (type 1; ~ 100 m)
10% of a decelerator (last 10 % most difficult)
6.25 GeV electron beam, 1.2 A
Nominal beam loading
CLICzero Parameters
Parameter Nominal value Unit
Drive Beam Energy 480 MeV
Pulse Length 6-140/ 243.7 ms / ns
Drive Beam Current (linac) 4.2 A
Decelerator Current 101 A
Combination Factor 24
Bunch Spacing 1.992 ns
Drive Beam Emittance ~100 mm mrad
Decelerator Bunch Length 1 mm
Repetition Rate 50 Hz
Main Beam Energy 6.5 GeV
Main Beam Current 1.2 A
Main Beam Pulse Length 156 ns
Main Beam Bunch Length ~0.5 mm
Main Beam Emittance ~30 mm mrad
Need for a global approach!Different solutions must be
explored (transformer based, fully solid state, HV & LV solutions)
CLIC studies & klystron modulators specs
Technology challenges
Reproducibility
Pulse to pulse reproducibility: 10 to 100ppm
Modulator and voltage measurement reproducibility never achieved before!
AC power quality optimizationMore than 1600 modulators pulsing synchronously! Utility grid power fluctuation minimized (~1%) – tough charger design
Machine availabilityWith more than 1600 modulators, reliability, modularity & redundancy must be optimized for maximum accelerator availability
Modulator topology selection considering:- Efficiency maximization (max. power
limited)- Reproducibility- Constant power consumption- Satisfactory accelerator availability
CLIC Drive Beam klystron modulators R&D strategy
What is needed to get
ready for CLICzero Prototyping and small series production of major hardware
DB-klystron, DB acc-structure, Modulator, Diagnostics, two beam modules
DB injector design and demonstration (Source, phase coding, stability) includes prototyping of DB rf system
Technical design of the DB linac Technical design of the beam combination complex
(delay loop, combiner rings) Technical design of the turn around loop Technical design of the probe beam injector Technical design and prototyping of the two beam modules
(acc-structure, PETS, diagnostics, Quads, stabilization and alignment)
Study and prepare location and implementation (tunnel, building) …..
Four cells structure with waveguide couplers
Hamed Shaker
In this design for the first SHB about 73 kW peak power is needed for 10ns filling time and 36.5 KV gap voltage.
Phase flipping simulation – 10 ns
CLIC Collaboration Working meeting- 2012
32Hamed Shaker
Excitation signal – port 1
Output signal – port 2
Output signal – port 1
≈ 14 ns
≈ 34 ns
Phase flipping simulation – 26 ns
CLIC Collaboration Working meeting- 2012
33Hamed Shaker
Excitation signal – port 1
Output signal – port 2
Output signal – port 1
≈ 20 ns
≈ 40 ns
Phase flipping- how fast?
CLIC Collaboration Working meeting- 2012
Borrowed from Oleksiy Kononenko
In the normal case the time interval between phase switching is constant (243.7 ns). In the Oleksiy model these intervals are not constant to have better energy dispersion at the end of main beam linac. It also give us an idea that how much the minimum phase flipping should be. The result shows us it should be less than 18ns. In my design I use 10ns similar to CTF3 SHBs.
34Hamed Shaker
Parameter for optimization
Travelling Wave Structure
For the known gap voltage and filling time our goal is to increase R/Q to reduce the input power. τ=10ns
V=36.5 KV
Pd : Power disappears on surface.β : Coupling coefficientQe = ωτ : External quality factorτ : Filling time
R : Effective shunt impedanceR’: Effective shunt impedance per length Q : Unloaded quality factorP: Source powerV : Gap voltageW’ : Stored energy per lengthL: Structure lengthvg : Group velocityn : Cell numbers
CLIC Collaboration Working meeting- 2012
35Hamed Shaker
Electron source options
CTF3: 1.6 ms, 9.6mC per pulse1 % droop specs7 nF, ~70 J stored energy
CLIC: 140 ms, 700 mC per pulse0.1 % droop specs 5 mF, ~50 kJ stored energy
gridded cathode might not survive
CTF3 gun concept might be not scalable for CLIC
Some simple considerations